Google将Gemini 2.5 Flash替换为更快捷、更便宜的Gemini 3 Flash in Search and app。 Google replaces Gemini 2.5 Flash with faster, cheaper Gemini 3 Flash in Search and app.

Google已推出Gemini 3 Flash(Gemini 3 Flash), 取代Gemini 2.5 Flash, Google has launched Gemini 3 Flash as the new default AI model in the Gemini app and Google Search, replacing Gemini 2.5 Flash. 其交付速度更快、效率更高,象征性品减少30%,每百万份投入品价值50美元,运行速度比以前版本快三倍。 It delivers faster, more efficient performance with 30% fewer tokens, costs $0.50 per million input tokens, and runs three times faster than previous versions. 该模型在推理、编码、数学和多式联运任务方面优异,优于早期版本,在SWE-Bench 验证和人类最后考试等基准中与较大的模型相匹敌。 The model excels in reasoning, coding, math, and multimodal tasks, outperforming earlier versions and rivaling larger models in benchmarks like SWE-bench Verified and Humanity’s Last Exam. 通过谷歌AI工作室、Vertex AI和其他平台,开发者可以进入全球免费获取该软件,从而增强日常的AI互动。 Available globally at no extra cost, it enhances everyday AI interactions and is accessible to developers via Google AI Studio, Vertex AI, and other platforms. 最新消息标志着谷歌在努力平衡速度、成本和情报方面迈出了重要一步。 The update marks a major step in Google’s effort to balance speed, cost, and intelligence.